From: bluhm <bluhm@openbsd.org>
Date: Sat, 6 Jan 2024 10:58:45 +0000 (+0000)
Subject: Take net lock before kernel lock.
X-Git-Url: http://artulab.com/gitweb/?a=commitdiff_plain;h=a61bf6f2a88e2df8799431eff1412c6579005e38;p=openbsd

Take net lock before kernel lock.

Doing KERNEL_LOCK() just before NET_LOCK() does not make sense.
Net lock is a rwlock that releases kernel lock during sleep.  To
avoid an unnecessary release and take kernel lock cycle, move
KERNEL_LOCK() after NET_LOCK().
There is no lock order reversal deadlock issue.  Both locks are
used in any order thoughout the kernel.  As NET_LOCK() releases the
kernel lock when it cannot take the lock immediately and has to
sleep, we always end in the order kernel lock before net lock after
sleeping.

OK sashan@
---

diff --git a/sys/net/if.c b/sys/net/if.c
index bb1b8919e92..8b9281d72aa 100644
--- a/sys/net/if.c
+++ b/sys/net/if.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: if.c,v 1.714 2023/12/29 11:43:04 bluhm Exp $	*/
+/*	$OpenBSD: if.c,v 1.715 2024/01/06 10:58:45 bluhm Exp $	*/
 /*	$NetBSD: if.c,v 1.35 1996/05/07 05:26:04 thorpej Exp $	*/
 
 /*
@@ -1774,16 +1774,16 @@ if_linkstate_task(void *xifidx)
 	unsigned int ifidx = (unsigned long)xifidx;
 	struct ifnet *ifp;
 
-	KERNEL_LOCK();
 	NET_LOCK();
+	KERNEL_LOCK();
 
 	ifp = if_get(ifidx);
 	if (ifp != NULL)
 		if_linkstate(ifp);
 	if_put(ifp);
 
-	NET_UNLOCK();
 	KERNEL_UNLOCK();
+	NET_UNLOCK();
 }
 
 void
diff --git a/sys/netinet/in.c b/sys/netinet/in.c
index d2207b3e124..7cf2cc6392c 100644
--- a/sys/netinet/in.c
+++ b/sys/netinet/in.c
@@ -1,4 +1,4 @@
-/*	$OpenBSD: in.c,v 1.185 2023/06/28 11:49:49 kn Exp $	*/
+/*	$OpenBSD: in.c,v 1.186 2024/01/06 10:58:45 bluhm Exp $	*/
 /*	$NetBSD: in.c,v 1.26 1996/02/13 23:41:39 christos Exp $	*/
 
 /*
@@ -260,8 +260,8 @@ in_ioctl(u_long cmd, caddr_t data, struct ifnet *ifp, int privileged)
 			return (error);
 	}
 
-	KERNEL_LOCK();
 	NET_LOCK();
+	KERNEL_LOCK();
 
 	TAILQ_FOREACH(ifa, &ifp->if_addrlist, ifa_list) {
 		if (ifa->ifa_addr->sa_family != AF_INET)
@@ -331,8 +331,8 @@ in_ioctl(u_long cmd, caddr_t data, struct ifnet *ifp, int privileged)
 		break;
 	}
 err:
-	NET_UNLOCK();
 	KERNEL_UNLOCK();
+	NET_UNLOCK();
 	return (error);
 }
 
@@ -353,8 +353,8 @@ in_ioctl_set_ifaddr(u_long cmd, caddr_t data, struct ifnet *ifp)
 	if (error)
 		return (error);
 
-	KERNEL_LOCK();
 	NET_LOCK();
+	KERNEL_LOCK();
 
 	TAILQ_FOREACH(ifa, &ifp->if_addrlist, ifa_list) {
 		if (ifa->ifa_addr->sa_family != AF_INET)
@@ -387,8 +387,8 @@ in_ioctl_set_ifaddr(u_long cmd, caddr_t data, struct ifnet *ifp)
 	if (!error)
 		if_addrhooks_run(ifp);
 
-	NET_UNLOCK();
 	KERNEL_UNLOCK();
+	NET_UNLOCK();
 	return error;
 }
 
@@ -409,8 +409,8 @@ in_ioctl_change_ifaddr(u_long cmd, caddr_t data, struct ifnet *ifp)
 			return (error);
 	}
 
-	KERNEL_LOCK();
 	NET_LOCK();
+	KERNEL_LOCK();
 
 	TAILQ_FOREACH(ifa, &ifp->if_addrlist, ifa_list) {
 		if (ifa->ifa_addr->sa_family != AF_INET)
@@ -527,8 +527,8 @@ in_ioctl_change_ifaddr(u_long cmd, caddr_t data, struct ifnet *ifp)
 		panic("%s: invalid ioctl %lu", __func__, cmd);
 	}
 
-	NET_UNLOCK();
 	KERNEL_UNLOCK();
+	NET_UNLOCK();
 	return (error);
 }